HPA-CRCE-032
The Measurement of X-Ray Beam Size from Dental
Panoramic Radiography Equipment
J R Holroyd
ABSTRACT
Dental panoramic X-ray equipment provides a challenge for the accurate measurement
of dose area product (DAP) values, which are utilised to help ensure radiation doses to
patients are kept as low as reasonably practicable.
This report describes a quick and accurate automated method using digitised images for
the measurement of the dimensions of panoramic X-ray beams. The accuracy of the
method is assessed by comparing its results against those obtained using a
transmission densitometer and micrometer to precisely measure the optical density
profile across radiographic images of panoramic X-ray beams. The method is also
compared to two established alternative methods: using a ruler alone and using a ruler
in combination with a light box and magnifying glass.
Both the digitised images method (average error of 7%) and the use of a ruler in
combination with a light box method (average error of 6%) were found to show good
agreement with the densitometer method. Comparing the ruler-only method to the
densitometer method showed that the ruler method underestimates the beam width by
on average 29%, but that this method could be accurately used for measuring the beam
height.
Digitised images can be used to measure panoramic beam profiles as accurately as
using a transmission densitometer, and with significantly improved speed compared to
both this method and use of a ruler with a light box and magnifying glass. The use of a
ruler alone is neither accurate nor consistent and should not be used to make these
measurements.
© Health Protection Agency
Centre for Radiation, Chemical and Environmental Hazards
Chilton, Didcot
Oxfordshire OX11 0RQ
Approval: April 2012
Publication: April 2012
£13.00
ISBN 978-0-85951-713-3
This report from the HPA Centre for Radiation, Chemical and Environmental Hazards reflects understanding and
evaluation of the current scientific evidence as presented and referenced in this document.
CONTENTS
1
Introduction
1
2
Method
2.1 Hardware and Software
2.2 Measurement method
2.3 Evaluation
2.3.1 Using a ruler
2.3.2 Using a ruler, light box and magnifying glass
2.3.3 Using a transmission densitometer
2
2
3
3
3
4
4
3
Results
5
4
Discussion
4.1 Beam width measurements
4.1.1 Comparison of the ruler method to the densitometer method
4.1.2 Reproducibility of the ruler method
4.1.3 Comparison of the light box and ruler method to the
densitometer method
4.1.4 Comparison of the digitised film method to the densitometer
method
4.2 Beam height measurements
12
13
5
Conclusion
14
6
References
14
8
9
9
10
11
iii
INTRODUCTION
1
INTRODUCTION
The Dental X-ray Protection Service (DXPS) of the Health Protection Agency (HPA)
provides an X-ray equipment performance testing service to the dental community
(Hewitt, 1984). The service employs ‘DXPS test packs’ that are used to remotely
assess the equipment performance of intra-oral and panoramic equipment.
A key component of this testing is to measure the radiation dose delivered to the patient.
National reference doses (NRDs) have been established for panoramic radiography
using the quantities, dose area product (DAP) (Institute of Physics and Engineering in
Medicine [IPEM], 2005) and dose width product (DWP) (Hart, Hillier and Wall, 2007).
DWP is the product of the radiation dose per exposure cycle and the width of the X-ray
beam profile and is usually expressed in the units of mGy mm. DAP is then the product
of DWP and the height of the X-ray beam, and is usually expressed in the units of
mGy cm2.
The DXPS panoramic assessment method requires a radiographic film (Kodak SR;
Kodak, Hemel Hempstead, England) to be exposed at the cassette carriage (or for
direct digital equipment, at the digital receptor) position for a complete radiographic
cycle during a standard adult panoramic exposure. The image captured on the
processed film can then be used to derive the dimensions of the X-ray beam at that
position. The radiation dose at the same position is measured separately using the
DXPS test pack. The dose area product can then be calculated and compared to the
national reference dose to check that the equipment is capable of restricting patient
doses to below the NRD providing proper exposure techniques are used. The
measurement of the beam width is not a straightforward task as the radiographic image
of the X-ray beam at the position of measurement is not sharply defined. In addition,
different models of panoramic equipment have significantly different beam sizes and
profiles. As typical measured beam widths are between 2 mm and 7 mm this
measurement is critically important in accurately calculating the DAP. A method that is
both accurate and reproducible is required. This true beam width is often not provided
by manufacturers of panoramic radiography equipment and when it is provided, the
method used to obtain the measurement is not supplied.
A number of methods have previously been proposed to measure panoramic beam
profiles with varying accuracies. One report indicated that using a ruler on an exposed
film gave a 20% overestimate of the beam width compared to measurement using an
optical density beam profile (Isoardi and Ropolo, 2003). Alternative methods involve the
use of in-beam ionisation chambers and DAP meters, to directly measure the dose area
product, (Williams and Montgomery, 2000; Tierris, Yakoumakis, Bramis and Georgiou,
2004) which have been shown to provide accurate measurements. However, these
methods require that the measurements are carried out onsite using specialist
equipment.
This paper presents an automated method utilising digitised beam images which is both
quick and easy to operate and has only a small associated cost. A similar method
utilising digitised images has been shown to agree within 8% on measurements of dose
width product compared to the direct measurement of dose using an ionisation chamber
1
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
(Doyle, Martin and Robertson, 2006) indicating that this may be a suitable methodology
for the measurement of panoramic X-ray beam profiles.
2
METHOD
2.1
Hardware and Software
A flat-bed scanner has been obtained (Epson Perfection V700 Photo; Epson (UK) Ltd,
Hemel Hempstead, England) which is capable of the transmission scanning of film
transparencies with an optical density range of up to 4D. Following darkroom
processing, the test pack films bearing the image of the X-ray beam are scanned using
the scanner’s default software package (Epson Scan; Epson (UK) Ltd, Hemel
Hempstead, England) and the resultant images are saved as 8-bit greyscale
uncompressed tagged image file format (tiff) files.
A custom software application has been designed using Microsoft Visual Basic 2005
(Microsoft; Redmond, USA). The image of a film is loaded into the program, the
greyscale values are read and converted to optical densities using a fourth order
polynomial relationship which was established for this type of X-ray film. The beam
width and height measurements can then be automatically computed, as illustrated in
Figure 1.
Figure 1 A typical beam film and the beam measurement software user interface
2
METHOD
2.2
Measurement method
The radiation dose delivered to the Kodak SR film can vary by a factor of ten for
different models of panoramic equipment. Due to this it is important that the X-ray beam
incident on the film is suitably filtered to ensure that the developed film is neither
saturated nor underexposed.
The DXPS test pack includes a number of copper filters of varying thicknesses. This
ensures that at least a portion of the beam profile will be clearly visible on the film for the
wide range of radiation exposures that the film may receive. For consistency, the beam
width is always measured at the position on the film where the peak optical density is
closest to 1. This ensures that the full beam profile is always used to calculate the beam
width, without the peak of the profile saturating the film or the film scanner.
To calculate the beam width, the area under the optical density profile is calculated and
this is divided by the peak optical density to give the width that would be expected of a
square profile with height equivalent to the peak optical density. Previous reports have
used a full width at half maximum (FWHM) measurement of the dose profile (Isoardi and
Ropolo, 2003; Doyle, Martin and Robertson, 2006) to determine the dose, however this
has been shown to underestimate the dose due to the shape of the dose profile
(Williams and Montgomery, 2000). Ten profiles are obtained across the filter area
selected and averaged to reduce inconsistencies associated with inhomogeneity of the
film. The beam height is a simple measurement of the length of the beam image. The
beam height is considerably greater than the width and typically has well defined edges,
therefore there is less chance of error on the measurement and any error has
significantly less influence on the dose area product calculation.
2.3
Evaluation
Twenty films were selected at random from a large selection of beam profile films. The
films were sorted into an order whereby similar beam profiles were not adjacent to each
other to reduce possible bias.
Three established alternative methods were also evaluated to assess their relative
ability to measure panoramic beam profiles: using a ruler (British Institute of Radiology
[BIR], 2001), using a ruler in combination with a light box and magnifying glass and
using a scanning densitometer (Isoardi and Ropolo, 2003). All the results were collected
using a standard data collection form and once all the measurements had been made
the results were collated and analysed.
2.3.1
Using a ruler
Four persons with experience of assessing panoramic equipment were selected to each
independently measure the beam width and height of the test films using only a ruler.
The reproducibility of using a ruler to measure the width of the beam profile was also
assessed. Two weeks after first measuring the films, two members of the department
were asked to repeat the measurements of the beam widths which were then compared
3
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
to their original measurements. The beam height measurements were not repeated as
there was considered to be far less subjectivity to this measurement.
2.3.2
Using a ruler, light box and magnifying glass
A different person measured the beam films using a ruler, but with the addition of a
magnifying glass and light box to assist with the measurements.
2.3.3
Using a transmission densitometer
Another person measured the width of the beams by manually obtaining an optical
density profile using a micrometer and a desktop transmission densitometer (Parry
DT1505; Alrad Instruments Ltd, Newbury, England) with a light source aperture of
0.2 mm. This method involved positioning the film on the densitometer and making
measurements of optical density as the film was gradually moved across the aperture in
increments of 0.2 mm. The resulting optical density profile could then be used to
accurately calculate the beam width using the same measurement method described
above which was used with the digitised images.
Figure 2 Transmission densitometer setup with a typical beam profile
Net OD (D)
1
0.8
0.6
0.4
0.2
0
0
4
1
2
3
4 5 6 7
Distance (mm)
8
9
10
RESULTS
RESULTS
The results of the measurements of the beam widths and heights of the 20 films, by all
four members of staff using only a ruler, are presented in Figures 3 and 4 below.
Figure 3 Mean and range of beam width measurements made using a ruler.
8
7
Beam width (mm)
6
5
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
Film number
Figure 4 Mean and range of beam height measurements made using a ruler.
180
170
160
Beam height (mm)
3
150
140
130
120
110
100
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20
Film num ber
The results of the repeat measurements of beam width, made two weeks after the initial
measurements, are shown in Table 1. It is apparent that the two members of staff
produced significantly different results from reading the same films, and with a differing
degree of consistency between their two individual reading sessions.
5
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
TABLE 1 Reproducibility of the ruler method when measuring beam width (measurements in mm)
Film
Number
Reader 3
st
1 Reading
2
nd
Reading
Reader 4
Difference
(%, rounded)
st
1 Reading
2
nd
Reading
Difference
(%, rounded)
1
4
4.5
13
6.5
6.5
0
2
3
3.5
17
4
4
0
3
4
4
0
7
6.5
-7
4
4
4.5
13
4.5
5
11
5
3.5
5
43
5
5
0
6
2
3
50
2
2
0
7
3.5
4.5
29
4.5
4.5
0
8
3
3.5
17
4.5
4.5
0
9
2
3.5
75
2.5
2.5
0
10
3
4
33
3.5
3
-14
11
4
5.5
38
5.5
5.5
0
12
2
3
50
2.5
2
-20
13
5
6.5
30
7.5
7.5
0
14
2.5
4.5
80
4
4
0
15
3
4
33
6
4.5
-25
16
2
2.5
25
2.5
2.5
0
17
3
3.5
17
6
5
-17
18
2.5
3
20
2.5
2.5
0
19
3.5
3.5
0
5
4
-20
20
4
4
0
5
5
0
Table 2 lists the results of the beam width measurements made using the remaining
three methods, while Table 3 lists the results of the beam height measurements made
using the ruler and light box method and the digitised image method.
6
RESULTS
TABLE 2 Beam width measurements (in mm) using a ruler and light box, densitometer or digitised
image. The digitised images were acquired using a scan resolution of 150 dpi.
Film number
Ruler and light box
Densitometer
Digitised image
1
6.0
6.1
6.2
2
4.5
4.6
4.9
3
6.5
6.2
6.3
4
5.0
5.3
5.6
5
5.5
5.4
5.8
6
2.5
2.8
3.0
7
5.0
4.7
5.0
8
5.0
5.0
5.5
9
3.0
2.8
3.0
10
4.0
3.7
4.1
11
5.5
5.9
6.4
12
3.0
3.2
3.4
13
8.0
7.7
8.0
14
4.5
4.4
4.8
15
5.0
5.8
5.8
16
3.5
3.2
3.4
17
6.0
6.2
6.7
18
3.0
3.1
3.3
19
5.0
4.8
5.1
20
5.5
5.9
6.1
7
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
TABLE 3 Beam height measurements (in mm) using both a ruler and light box and a digitised
image
Film number
4
Ruler and light box
Digitised Image
1
155
155
2
114
113
3
156
155
4
129
129
5
120
119
6
106
106
7
117
118
8
150
148
9
127
126
10
131
131
11
134
134
12
132
131
13
162
162
14
138
138
15
140
140
16
126
125
17
140
140
18
133
132
19
152
150
20
134
132
DISCUSSION
The beam width results obtained using the densitometer method were taken to be the
“true” result as this method has been shown to be an accurate means for making this
measurement (Isoardi and Ropolo, 2003). In the discussion which follows, the
measurement results obtained using the alternative methods are, therefore, presented
as a percentage deviation compared to the value obtained using the densitometer. For
the beam height measurements, the digitised image method was considered to be the
accurate method as the densitometer was not used for this assessment due to the
labour intensive nature of the method and the reasonable expectation that this
measurement can be made accurately using simpler methods.
8
DISCUSSION
4.1
Beam width measurements
4.1.1
Comparison of the ruler method to the densitometer method
TABLE 4 Percentage error on measuring beam width using a ruler
Film Number
Reader 1
Reader 2
Reader 3
Reader 4
Average
Min
Max
1
-50
16
-34
7
-15
-50
16
2
-34
-12
-34
-12
-23
-34
-12
3
-51
-11
-35
14
-21
-51
14
4
-16
-16
-25
-16
-18
-25
-16
5
-35
-35
-35
-7
-28
-35
-7
6
-29
-64
-29
-29
-38
-64
-29
7
-47
-26
-26
-4
-26
-47
-4
8
-40
10
-40
-10
-20
-40
10
9
-46
-64
-29
-11
-38
-64
-11
10
-47
-73
-20
-7
-37
-73
-7
11
-33
-33
-33
-8
-26
-33
-8
12
-52
-68
-37
-21
-44
-68
-21
13
-23
-29
-35
-3
-23
-35
-3
14
-43
-43
-43
-9
-35
-43
-9
15
-56
6
-47
6
-23
-56
6
16
-53
-69
-38
-22
-45
-69
-22
17
-18
-2
-51
-2
-18
-51
-2
18
-34
-34
-18
-18
-26
-34
-18
19
-37
-16
-26
5
-18
-37
5
20
-32
-40
-32
-15
-29
-40
-15
RMSE*
40
40
34
13
29
50
13
Standard
Deviation
12
28
8
11
9
14
12
* RMSE is the root mean squared error
The values measured with a ruler underestimate the beam width by an average of 29%.
However, there is a large variation between one reader (Reader 4) and the other three.
Reader 1, Reader 2 and Reader 3 underestimate the beam width by an average of 40%,
40% and 34% respectively whereas Reader 4 underestimates the beam width by an
average of 13%. For five films (films 1, 3, 8, 15 and 19) one or more readers measured
the beam width to be greater than the measurement made using the densitometer
method. These five films also show a wide range of measurements between the four
members of staff (for film 1 the range of measurements is 4 mm or 65%, see Table 4).
On examining the beam width profiles it can be seen that these beams are particularly
diffuse which makes it difficult to determine the edges of the beam and therefore the
measurement is highly subjective to the individual making the measurement, as
illustrated in Figure 5. For comparison, film 4 has well defined beam edges and this film
has been measured within 11% (0.5 mm) by all readers as can be seen in Figure 6.
9
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
Figure 5 Image of film 1 which has a very diffuse edge and the corresponding beam profile
determined by the software
Figure 6 Image of film 4 which has a very sharp edge and the corresponding beam profile
determined by the software
4.1.2
Reproducibility of the ruler method
The absolute measurement difference between the 1st and 2nd beam width
measurements of the twenty test films are presented in table 5. Reader 4 shows a high
correlation between repeated measurements, with only 3 out of 20 films having a
difference in measurement of 1 mm or greater. Reader 3 has a difference of 1 mm or
greater in reading 10 out of 20 films. The difference between measurements
demonstrated the subjectivity of the ruler method.
The differences in beam
measurements between readers are considerable and these results demonstrate that
the differences between successive measurements of a film by an individual are also
significant.
10
DISCUSSION
TABLE 5 Absolute difference (in mm) between 1st and 2nd beam width measurements
Film Number
Reader 3
Reader 4
1
-0.5
0.0
2
-0.5
0.0
3
0
0.5
4
-0.5
-0.5
5
-1.5
0.0
6
-1
0.0
7
-1
0.0
8
-0.5
0.0
9
-1.5
0.0
10
-1
0.5
11
-1.5
0.0
12
-1
0.5
13
-1.5
0.0
14
-2
0.0
15
-1
1.5
16
-0.5
0.0
17
-0.5
1.0
18
-0.5
0.0
19
0
1.0
20
0
0.0
RMSE
1.0
0.5
Standard Deviation
0.6
0.5
4.1.3
Comparison of the light box and ruler method to the densitometer
method
The differences between using the densitometer method and the light box and ruler
method are presented in table 6. The RMSE error is seen to be 6% which shows good
correlation between the two methods. The use of a light box and ruler is still a
subjective measurement and to maintain good quality results would require the
individuals performing the measurements to regularly check their results against
standard films whose beam widths had been determined by a method similar to the
densitometer method. A further experiment may be for several people to repeat the
measurements using the light box method to establish whether this is an appropriate
method, and can demonstrate good repeatability, or whether the good results observed
in this experiment are due to the individual who carried out the measurements.
11
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
TABLE 6 Percentage difference between
the densitometer method and using the
ruler in combination with a light box
Film Number
% Difference
1
-2
2
-2
3
5
4
-6
5
2
6
-11
7
6
8
0
9
7
10
8
11
-7
12
-6
13
4
14
2
15
-14
16
9
17
-3
18
-3
19
4
20
-7
RMSE
6
Standard Deviation
6
4.1.4
Comparison of the digitised film method to the densitometer method
The differences between using the densitometer method and digitised images method
are presented in table 7. The RMSE error is seen to be 7% which shows good
correlation between the two methods. The width measured by the digitised image
method is generally greater than that read using the densitometer method. The digitised
image method uses an average of ten rows of data to account for small differences in
the properties of the film. As the densitometer calculates an optical density for a single
row of data it is more likely to be susceptible to small changes in the film and this may
account for the differences seen in the width measurements.
12
DISCUSSION
TABLE 7 Percentage difference between
the densitometer method and the digitised
image method
4.2
Film Number
% Difference
1
2
2
7
3
2
4
6
5
7
6
7
7
6
8
10
9
7
10
11
11
8
12
6
13
4
14
9
15
0
16
6
17
8
18
6
19
6
20
3
RMSE
7
Standard Deviation
3
Beam height measurements
Panoramic X-ray beam heights are typically around 120 mm or 150 mm corresponding
to the two heights of film cassette commonly used. As such it would be expected that
this distance can be accurately measured using a ruler graduated in 1 mm steps.
Additionally, the beam profile has well defined start and end points that can readily be
measured between. The error in measuring the beam height was an RMSE of
approximately 1% which indicates that both the digitised images method and the ruler
method could be utilised for measuring panoramic beam heights.
13
THE MEASUREMENT OF X-RAY BEAM SIZE FROM DENTAL PANORAMIC RADIOGRAPHY EQUIPMENT
5
CONCLUSION
The height of the X-ray beam can be accurately measured using all of the methods
evaluated in this report. The simplest method, using only a ruler, can be adopted as the
method of choice for this measurement. However, using a ruler alone cannot be
considered a satisfactory method for the determination of panoramic X-ray beam width
due to both the inaccuracy of the method and the subjectivity that has been clearly
demonstrated in this report, both between different persons and the same people at
different points in time. The use of a ruler, light box and magnifying glass may be
considered an acceptable method, although if this method was to be adopted further
evaluation should be carried out to ensure accuracy and consistency. This report has
shown that the use of digitised images and transmission densitometry are satisfactory
methods for the measurement of panoramic beam widths. DXPS has adopted the use
of digitised images as this method offers a simpler and quicker automated measurement
than the manual use of a transmission densitometer.
6
REFERENCES
British Institute of Radiology (BIR), 2001. Assurance of quality in the diagnostic imaging department
(2nd Ed). London: BIR.
Doyle, P., Martin C.J., Robertson, J., 2006. Techniques for the measurement of dose width product in
panoramic dental radiography. Br. J. Radio., 79, pp.142-7.
Hart, D., Hillier, M.C., Wall B.F., 2007. Doses to patients from Radiographic and Fluroscopic X-ray
Imaging Procedures in the UK – 2005 Review. HPA-RPD-029. Chilton: Health Protection Agency.
Hewitt, J.M., 1984. The development and Operation of a Method for the Remote Determination of Xray beam Parameters Used in Dental Radiography. NRPB-R164. London: HMSO.
Institute of Physics and Engineering in Medicine (IPEM), 2005. Dental Radiography. In: Report 91:
Recommended Standards for the Routine Performance Testing of Diagnostic X-ray Imaging
Systems. Ch.9. York: IPEM.
Isoardi, P., Ropolo, R., 2003. Measurement of dose-width product in panoramic dental radiology, Br. J.
Radiol., 76, pp.129-131.
Tierris, C.E., Yakoumakis, E.N., Bramis, G.N., Georgiou, E., 2004. Dose area product reference levels
in dental panoramic radiology. Radiat. Prot. Dosim., 111, pp.283-7.
Williams, J.R., Montgomery, A., 2000. Measurements of dose in panoramic dental radiography, Br. J.
Radiol., 73, pp.1002-6.
14